Diane Coyle: "In the end it will be people in general who benefit from artificial intelligence"

The economic effects of artificial intelligence are beginning to be felt in specific tasks, but not yet in aggregate productivity figures. Diane Coyle, Bennett Professor of Public Policy at the University of Cambridge and one of the confirmed speakers at La Toja Forum 2026, believes the explanation lies with companies: before they can reap the benefits, they must reorganise their workflows and business models. Nor does she share the most extreme forecasts: "I don't myself think it will be a job apocalypse of the kind that some of the AI leaders have been talking about".

In conversation with Agenda Pública, Coyle discusses the technology's impact on youth employment, the distribution of productivity gains and Europe's gap with the United States. For the continent, the opportunities lie in data, applications and research, although technological dependence creates a geopolitical vulnerability. The decisive question, she warns, will be who captures the value created: "We are in a situation where a very small number of people are making a lot of money".
  Do you think we are overestimating the economic impact of AI, or is it simply that we have not seen the real effects yet?

It's a bit of both. It depends what you mean by AI. There is a lot of hype around artificial general intelligence and conscious AI. I think those things are complete hype. If you think about AI tools as an evolution of machine learning or earlier digital tools, then I think it's probably a question of it being too early to see what the effects are.

There are a lot of studies now showing that, for individual tasks at work, you save a lot of time using AI. Individual people can save time and do their work better, but that is not yet translating into organisational productivity gains. And that, to me, says we need to see the kind of reorganisation of workflows and new kinds of business models that we saw in the late 2000s when digital platforms came along.

"We need to see the kind of reorganisation of workflows and new kinds of business models that we saw in the late 2000s when digital platforms came along"

If that's right, then in the next two or three years we should start seeing measurable macro-level productivity gains from AI. But that kind of organisational change is expensive, it's slow and people don't like it, so it always takes time. It's much slower to change institutions than to change the technology.

That was actually my next question. Are we asking the wrong question when we focus on AI destroying jobs? Should we be paying more attention to who will capture the productivity gains that AI creates?

I think I agree with that. The evidence about young people's jobs is a bit tentative because there are business-cycle changes that you need to try to separate out. There was a lot of hiring and retention post-COVID, and then there's a sort of natural downturn in entry-level jobs.

It is the case that the kinds of companies that hired lots of young people, such as accountancy firms or law firms, have cut back on that kind of hiring, but not entirely, because they understand that you need young people in order to have experienced and senior people over time. So they are worried about those trajectories and about how people gain experience if you're not hiring young people.

Having said that, the character of jobs will change for sure and there will be some disruption. I've been thinking quite a lot about legal services because of the evidence from Google's Atlas study and other tech companies that there's a lot of use of AI in the household, and one of the things people are using it for is administration.

Two friends of mine recently had to get probate for their parents' wills. One paid £12,000 using a lawyer, and the other got it for free using ChatGPT. So that says there is an opportunity for some kind of disintermediation. Again, we saw this in previous waves of digitalisation with online banking, online taxis, Uber, Airbnb and so on.

"Two friends of mine recently had to get probate for their parents' wills. One paid £12,000 using a lawyer, and the other got it for free using ChatGPT"

All of those disruptive models did have impacts on real-world sectors of the economy. So I think we will see some of that kind of disruption. I don't myself think it will be a job apocalypse of the kind that some of the AI leaders have been talking about. I think they're exaggerating.

And what could governments do to make sure that workers gain something from this?

We know from previous waves of automation that governments, on the whole, are not very good at helping people navigate those transitions. Some are better. People always point to Denmark's flexicurity model, which provides generous support for people as they learn new skills and transition to new jobs.

So we can learn from past policy mistakes. And the other set of questions is about who gets the value created by using data that we all produce and which is captured by companies. Or, if my employer makes a digital twin of me and gets one and a half times as much work, what part of that do I get in a pay increase?

I think there are questions about allocation that go back to some fundamentals of countervailing power in the labour market: what bargaining power do workers have and to what extent is a government going to support them in capturing some of that value when it manifests itself in productivity gains?

We've spent years talking about Europe's widening productivity gap with the United States. But how much of that diagnosis is actually about technology itself, and how much is about investment, scale and other structural factors?

I think a lot of it is about those kinds of structures: regulatory barriers that make it difficult to get into new markets.

If you think about why Airbnb or Uber did so well, it's because there was regulatory arbitrage. Taxi markets were highly regulated and expensive. There was an open goal there for a tech company to come in.

"If you think about why Airbnb or Uber did so well, it's because there was regulatory arbitrage. Taxi markets were highly regulated and expensive"

There are also things like the cost of laying off workers. Obviously, we want to protect people, but the costs are excessive in some countries. It varies quite a lot. In Scandinavia, that's not the case at all, and they are the most dynamic and high-tech economies in Europe.

Whereas if you look at Germany or France, by contrast, it's extremely expensive to set up a start-up and then have it fail a year later, which is what happens with tech start-ups. The costs that you incur to do that are very high indeed.

So the return venture capitalists need in order to invest in those kinds of start-ups is several percentage points higher than they would need in the United States or in Scandinavia.
Is Europe at risk of becoming a very sophisticated customer of other countries technologically, rather than developing genuine technological capacity of its own?

Well, it's not a bad thing to import services from other countries if you have services of your own that you can export. And there are also the Chinese models, which offer, for most applications, something just as good as the American models. So there is some competition among the frontier AI models.

"It's not a bad thing to import services from other countries if you have services of your own that you can export"

European countries therefore need to think about what advantages they have. An obvious one is data and applications. Actually, this has been the Chinese path as well. They have concentrated at least as much on applications of AI as on the frontier models themselves.

So if we have great applications or, as we do in Europe, fantastic AI researchers, then that gives us some stake in a positive global AI economy.

I think the worry here is about geopolitics more than economics. To what extent does an unpredictable American administration, which has legal power over what the AI companies can do, create vulnerabilities or not?

There is that familiar criticism that the United States innovates, China scales and Europe regulates. Do you think that diagnosis is fair, or does it confuse regulation with other problems we may have, such as lack of investment and market fragmentation?

I think it's a bit unfair. Both the US and China have the huge advantage of scale. The venture capital market and the scale of the domestic market make a big difference.

Europe could do more to bring down barriers to either production or demand, so that there are more opportunities to scale. But I think we are in danger of talking ourselves down too much in Europe. There are genuine research advantages.

You can imagine that a European university is going to be the place where much more energy-efficient technologies are developed, for example, and that becomes a great competitive advantage. This is a really dynamic area. The research frontier moves very quickly.

There are lots of opportunities. So I guess the challenge is figuring out which barriers entrepreneurial companies cannot get around. I would focus on things like the ability to use data to innovate and other regulatory barriers that mean legal technology or health technology companies and start-ups can't grow.

And I think all of us, including the UK, have that kind of barrier that we need to break down.

If you had to bet today on who will capture the largest share of the economic gains from artificial intelligence —workers, companies or governments— where would you put your money?

It's a good question. I think eventually it will have to be people. All these fundamental general-purpose technologies eventually deliver benefits to people in general, because that's the market.

You've got to have demand for those technologies to be used. The alternative —and sometimes it happens on the way— is revolution. We are in a situation where a very small number of people are making a lot of money.

Unless that value starts being shared, we're going to be in very politically disrupted times. So I don't know whether we'll take the smooth route or the stormy route, but I think in the end it will be people in general who benefit from artificial intelligence.

Thank you so much. See you at La Toja.